Age | Commit message (Collapse) | Author |
|
Signed-off-by: Munyoki Kilyungi <me@bonfacemunyoki.com>
|
|
Signed-off-by: Munyoki Kilyungi <me@bonfacemunyoki.com>
|
|
* gn3/db/rdf.py (sparql_query): Parse CONSTRUCTS and SELECTS
differently.
(strip_url): Rename to ...
(get_url_local_name): ... this.
(get_dataset_metadata): Fetch extra fields in RDF.
Signed-off-by: Munyoki Kilyungi <me@bonfacemunyoki.com>
|
|
Signed-off-by: Munyoki Kilyungi <me@bonfacemunyoki.com>
|
|
Signed-off-by: Munyoki Kilyungi <me@bonfacemunyoki.com>
|
|
In certain scenarios (probably due to host contamination) the effective UID of
the running application is not the expected "genenetwork" user, rather, it is
a user on the host system, which leads to the error:
KeyError: 'getpwuid(): uid not found: 1000'
This commit prevents the application from failing in such a case, but still
logs out such weirdness.
|
|
|
|
Useful for debugging issues within the guix containers.
|
|
This information is useful to assist with debugging issues within the guix
containers.
|
|
|
|
* gn3/db/rdf.py: Importh pymonad.Nothing.
(get_dataset_metadata): Replace SELECT with CONSTRUCT.
Signed-off-by: Munyoki Kilyungi <me@bonfacemunyoki.com>
|
|
* gn3/db/rdf.py: Import unquote and urlparse.
(strip_url): New function.
Signed-off-by: Munyoki Kilyungi <me@bonfacemunyoki.com>
|
|
|
|
Add delays to avoid overwhelming the DB server.
|
|
|
|
* scripts/index-genenetwork (main): Write table checksums into index.
|
|
* scripts/index-genenetwork (SQLTableClause): New variable.
(genes_query, phenotypes_query): Express tables using SQLTableClause.
(serialize_sql): Serialize SQLTableClause.
|
|
* scripts/index-genenetwork (write_document, index_query): Fold long lines.
|
|
* scripts/index-genenetwork (main): Ensure no other indexing job is running.
|
|
* scripts/register_sys_admin.py: fix linting error
* tests/unit/auth/fixtures/oauth2_client_fixtures.py: hash client secret in
database. The code expects it hashed in the database.
|
|
|
|
To ease registration of system-admin user, provide a CLI script to register
the user and mark them as sys admin in one go.
|
|
We are saving the client secret in an encrypted form, meaning we have to
verify that the CLIENT_SECRET that is provided is the same one as was
generated at registration in a different way.
Initially, I was doing a direct comparison, having saved the CLIENT_SECRET
value as unencrypted plain-text.
|
|
Add UI and code to enable the administrative user to register new OAuth2
clients that can access the API server.
|
|
* gn3/auth/authentication/oauth2/views.py: Remove endpoint
* gn3/auth/authorisation/users/admin/__init__.py: New admin module
* gn3/auth/authorisation/users/admin/ui.py: New admin module
* gn3/auth/authorisation/users/admin/views.py: New admin module
* gn3/auth/views.py: Use new admin module
* gn3/errors.py: Fix linting errors
* gn3/templates/login.html: New html template
* main.py: Fix linting errors
|
|
|
|
* docs/authentication_and_authorisation/oauth2_clients.md: New documentation
* gn3/auth/authentication/oauth2/server.py: Raise appropriate error if no
client is found.
* gn3/auth/authentication/oauth2/views.py: Handle exception in the case where
a UI should be presented to the user, rather than presenting the raw JSON
response to the user.
* gn3/errors.py: Handle any authlib OAuth2Error at the application's top-level
* gn3/templates/oauth2/oauth2_error.html: Handle any authlib OAuth2Error at
the application's top-level
|
|
|
|
Document some CLI utility commands useful for development and for supporting
the operation of the GN3 app.
|
|
|
|
A script to assign existing data not assigned to any group to publicly-visible
resources.
|
|
|
|
Enable users with the appropriate privileges to masquerade as other users by
providing an endpoint that provides a new token for the "masqueradee"
|
|
|
|
|
|
|
|
|
|
|
|
Also set default value for os.environ.get("GENENETWORK_FILES")
|
|
|
|
initialized (if the genofile was empty, for example)
|
|
Initialize samplelist variable
|
|
|
|
This is necessary in order to allow for editing the values of samples that don't currently have values
|
|
There's probably a better way to fix this query (it was previously returning each sample twice), but DISTINCT was the easiest way I could come up with
|
|
|
|
|
|
after the recent changes to the DB
|
|
|
|
|